1. Field of the Invention
This invention relates to a communication control device for carrying out serial data transmission and reception with a communication network, and particularly to a communication control device which can realize high speed operation in a high-speed local area network.
2. Description of the Prior Art
Recently, high-speed data transmission in an LAN (Local Area Network) has been studied. Particularly, the high-speed data transmission in the LAN by the so-called token ring method which is standardized in IEEE 802.5 is now noted.
As communication data used in the token ring method, the so-called differential Manchester codes (DM codes) are used. In the system of the codes, a time interval which is required for transferring data as a unit of information is divided into two parts as shown in FIG. 1. The signal level of the former half part of the time interval is compared with that of the latter half part of the previous time interval thereof, and the signal level of the latter half part of the time interval is compared with that of the former half part thereof. In FIG. 1, four states of combination of inversion and non-inversion of the former half part and the latter half part of the time interval are shown with respect to each state (H level or L level) of the latter half part of the previous time interval. These states are called "J", "K", "1", "0" respectively as shown in FIG. 1. In this case, when inversion is denoted by "1" and non-inversion is denoted by "0", the DM codes "J", "K", "1", "0" are expressed as "11", "01", "10", "00" respectively. These new codes respectively comprising a combination of the two binary bits are called ER codes. Herein, the lower bit of each of the ER codes is called a code violation bit, and the higher bit thereof is called a data bit.
In communication data in which each unit of information is serial as shown in FIG. 1, no discriminating factor for showing a boundary between serial two units of information is included in each of the units. Therefore, it is very likely that two bits comprising the latter half part of the previous unit of information and the former half part of a unit of information are recognized to be one unit of information. To prevent this problem, a start code of data or a discrimination code of boundary expressed by "JKOJKOOO" comprising the DM codes is used as a start delimiter (SD), and the units of information starts from this code by two bits.
For example, as shown in FIG. 2, there is a terminal of a conventional LAN system based on the token ring method in which such DM codes are used.
In FIG. 2, a media interface 1 is mainly composed of an analog circuit, and regenerates reception data having some logic level from a weak input signal given by a network, and also extracts a reception clock signal overlapping in the input signal. The reception data regenerated by the media interface 1 are converted into reception information by a network control device 2 in synchronism with the reception clock signal extracted, then the reception information is given to a data processing device 3.
Moreover, the data processing device 3 generates transmission information. Then, the transmission information is given to the network control device 2, and converted into transmission data thereby. The transmission data are further converted into an analog output signal by the media interface 1, thereafter the output signal is outputted to the network, and transmitted therefrom.
In such construction, the communication data are transferred on the network as bits arranged in series, and transmitted or received through the media interface 1. Therefore, between the media interface 1 and the network control device 2, the transmission data or the reception data are serially transferred by one bit through a transmission data line 4 or a reception data line 5. Moreover, through another reception clock line 6, the reception clock signal is given to the network control device 2 from the media interface 1.
Accordingly, one unit of information of transmission data expressed by the DM code as described above is processed as the ER code comprising two bits, and the ER code is serially transferred by one bit through the transmission data line 4 or the reception data line 5 so as to be processed by one bit. Namely, the transmission data or the reception data must be recognized and processed by the media interface 1 and the network control device 2 not by one DM code but by one bit. However, the reception signal must have a frequency which is synchronous to one bit and is twice as high as the transfer speed of one unit of information.
As mentioned above, in the conventional LAN system based on the token ring method in which the DM code is used, it is necessary that the clock signal has a frequency twice as high as the transfer speed of communication data.
Therefore, when the transfer speed of the communication data becomes higher, a circuit included in the system for generating the clock signal must generate a clock signal having a higher frequency, so that the circuit must be operated at a higher speed. However, when the operation speed becomes high, the power consumption is inevitably increased. Moreover, by the increase of power consumption, a large-scale power source and system construction capable of heat radiation must be required. Accordingly, it is difficult to realize a small-scale system by the conventional method.
Moreover, to realize the high-speed operation, it is necessary to compose the circuit more finely. Moreover, the delay time on transmitting signals must be more reduced. As the result, the circuit design and layout design for the communication control device and the manufacture thereof become very difficult.